Self-Directed Learning in Problem-Based Learning Versus Traditional Lecture-Based Learning: A Meta-Analysis

Problem-based learning is an approach to education and learning that focuses on students investigating problems, rather than being directly instructed by teachers. Reviews, also called meta-analyses, comparing traditional lecture-based learning to problem-based learning report modest positive learning gains in student cognitive outcomes. Many meta-analyses have been conducted to analyze the effectiveness of problem-based learning, but none examine the extent of self-directed learning in problem-based learning. The purpose of this study was to conduct a meta-analysis across all disciplines (medicine, education, business, history, etc.) while examining self-directed learning in problem-based learning. This study used a random effects model meta-analysis using 75 outcomes from 38 studies. A test of heterogeneity indicated genuine variance across outcomes (Q = 559.57, df = 74, p \u3c 0.01), supporting the use of a random effects model. Results indicated a statistically significant overall medium effect size, g = 0.45, z(74) = 7.11, p = 0.01, favoring problem-based learning over traditional lecture-based learning, indicating an expert is likely to detect differences through casual observation while a nonexpert might see them is if looking closely. Subgroup analyses indicate positive effect sizes for the four components of self-directed learning with two being statistically significant. Findings and conclusions provided the first synthesis of noncognitive outcomes in problem-based learning by specifically analyzing self-directed learning. From this synthesis, practitioners learn that problem-based learning promotes noncognitive skills in self-directed learning